CarGurus Inc (CARG) 2018 Q2 法說會逐字稿

Greetings, and welcome to the CarGurus Second Quarter 2018 Earnings Results Conference Call. (Operator Instructions) As a reminder, this conference is being recorded. I would now like to turn the conference over to our host, Rodney Nelson, Investor Relations for CarGurus. Thank you. You may begin.

Thank you, operator. Good afternoon, and welcome to CarGurus' Second Quarter 2018 Earnings Call. We'll be discussing the results announced in our press release issued today after the market closed and posted on our Investor Relations website.

With me on the call today is Langley Steinert, CarGurus' Founder and Chief Executive Officer; Jason Trevisan, Chief Financial Officer; and Sam Zales, Chief Operating Officer.

During the call, we will make statements related to our business that may be considered forward looking, including statements concerning our financial guidance for the third quarter and full year 2018, our business and growth strategy and our plans to execute on our growth strategy, our brand awareness efforts, our investments in and ability to drive adoption of new and existing products, the growth levers we expect to drive our business, our ability to maintain existing and acquire new customers, our expansion into international markets, the expected contributions of new addition to our Board of Directors, the impact on our cash flows from tax withholding payments related to net share settlements in future periods and other statements regarding our plans, prospects and expectations.

Forward-looking statements may include words and phrases such as we expect, we believe, we intend, we anticipate, we plan, may, upcoming and similar terms. These statements reflect our views only as of today and should not be considered our view as of any subsequent date. We undertake no obligation to update or revise these forward-looking statements except as required by law. Forward-looking statements are not promises or guarantees of future performance and are subject to a variety of risks and uncertainties that could cause the actual results to differ materially from our expectations. For a discussion of material risks and other important factors that could affect our actual results, please refer to those contained under the heading Risk Factors in our quarterly report on Form 10-Q filed after today's market close as may be updated by our other SEC filings.

Further, during the course of today's call, we will refer to certain non-GAAP financial measures. A reconciliation of GAAP to non-GAAP measures is included in our press release issued after market close today. The press release and our SEC filings can be found on the Investor Relations section of our website at investors.cargurus.com and the SEC's website at sec.gov.

With that, I'll turn it over to Langley.

Thanks, Rodney, and thanks to everyone for joining us today. Before I begin, I want to welcome Rodney Nelson to the CarGurus team. Rodney will be running our Investor Relations effort, and we're excited to have him onboard.

We enjoyed a strong second quarter as we continue to execute against our strategic initiatives for 2018. Our success continues to be driven by our commitment to providing consumers with the best possible car shopping experience. Our transparent, technology-driven platform and our investments in our brand continue to fuel growth in our market leading audience.

The high-quality engagement we generate from our audience ultimately yields robust ROI for dealers, and we are seeking to further bolster dealer ROI with innovation in both existing and new dealer products. As a result, we delivered 45% revenue growth in the second quarter, while exceeding our Q2 operating income guidance. Based on the strength of these results and our continued execution, we are increasing our guidance for the full year, which Jason will detail in a few moments.

CarGurus is a data-driven company. We work tirelessly to continually improve our platform and bring order to the millions of unstructured data points across the automotive market. Our technology-driven approach is clearly resonating with our large and highly engaged user base. The powerful combination of robust vehicle information, proprietary deal ratings and insightful dealer reviews is attracting more car shoppers to CarGurus than any other of our competitors in the U.S.

Average monthly unique visitors to our U.S. marketplace increased 56% year-over-year in the second quarter. This was an acceleration from 33% growth in the first quarter of 2018. In Q2, our average monthly U.S. user sessions also accelerated to 52% growth year-over-year, up from 37% session growth in Q1.

As we proudly shared with you last quarter, CarGurus is now the leading U.S. automotive marketplace as measured by comScore unique monthly visitors and minutes spent on our site. The investments we've begun making in our brand are bearing fruit, evidenced by our growth in aided and unaided awareness among car shoppers as well as disproportionate growth in our direct traffic and branded search volumes.

We are also using targeted social media campaigns to increase our aided awareness amongst consumers that have not been to our site before, and we're seeing those investments reflected in our traffic growth. These results from our brand investments are a key driver in the acceleration in audience growth we witnessed in the second quarter. In fact, in the second quarter, CarGurus had 60% more unique visitors to our U.S. site than our nearest major automotive marketplace competitor, according to comScore, which we believe is a testament to our consumer-centric approach. We are winning consumers with a powerful combination of product, user experience, algorithmic traffic acquisition and now brand building; and as a result, we are growing our audience lead.

Not only are we attracting more car shoppers to our platform than our major competitors, but our audience spent more time on our site than the audience of the next 3 major U.S. automotive marketplaces combined. We believe our extensible platform plays a key role in this dynamic, evidenced by our mobile traffic growth. We are the clear leader among U.S. competitors in mobile unique monthly visitors as measured by comScore. And we now generate nearly 80% more mobile visitors than our next closest major U.S. competitor. Our progress in audience growth, mobile engagement and brand awareness is also highly appreciated by our paying dealers, who truly value partnering with a large, high-quality and well-known marketplace.

We are succeeding in disrupting our domestic market, but our work is far from over. In order to maintain our trajectory, we are laser focused on providing a platform that delivers the best possible experience for consumers with our relentless commitment to earning trust and delivering transparency. Moving forward, we aim to further extend our #1 position in the U.S. while we build our market position in other countries around the world.

To that extent, we continue to make strides in our international audience, where we generated year-over-year unique monthly visitor and session growth of 54% and 60%, respectively, in the second quarter. In total, global average monthly user sessions increased 53% year-over-year, topping 100 million average monthly sessions for the first time. As a reminder, we evaluate potential new markets on a country-by-country basis in order to understand consumer buying behaviors, dealer practices and the competitive landscape.

Last quarter, we outlined that we will report on new country launches once our site emerges from its beta launch. With that said, I'm pleased to announce that in the second quarter, we launched our disruptive marketplace in Spain. We believe that Spain has many of the same market dynamics that have made our focus on transparency and unbiased information successful elsewhere. Needless to say, it is early days in other Spanish business, but we are excited about our opportunity in that region.

Elsewhere in our international segment, momentum continues to build in the U.K., where we were, again, the fastest growing major automotive shopping site by unique monthly visitors in Q2 as measured by comScore. Our average unique monthly visitors in the U.K. grew 58% year-over-year in Q2, and dealers are taking notice. The Jardine Motors Group, one of the U.K.'s largest dealer groups, has seen a substantial increase in consumer engagement since joining our platform.

Alex Brown, Jardine's Head of digital, noted that since becoming a paying dealer last year, Jardine's conversion rates, time on site, phone call durations and other engagement metrics show that the quality of CarGurus' interactions is at the upper end of advertising partners. Brown adds, "What sparked our interest in CarGurus was that we knew disruption was coming to this market, and we thought CarGurus would be a strong partner given their success in the U.S."

Dealers are recognizing the value of our engaged audience and data-driven marketplace, and our product innovation strategy is extending that value beyond listings. The primary strategic initiative for us this year is introducing new products and driving dealer adoption. Our first dealer advertising product, Dealer Display, continues to perform well as it gives U.S. dealers the opportunity to get more exposure to the tens of millions of visitors on our site each month. We are also seeing encouraging traction with our dealer search engine marketing product, SEM Plus. After years of optimizing our own programmatic SEM strategy across billions of key words to drive traffic to our marketplace, we are now leveraging our domain expertise on behalf of our dealers for their own websites. Search engine marketing is a product that dealers have been inquiring about for some time, but they simply do not have the ability to effectively invest in this practice at scale on their own. And the small local agencies lack the technology investment to maximize the dealer's search marketing potential.

As SEM Plus completes its second full quarter out of beta, we're receiving great feedback from customers. Colorado-based King Buick GMC and King Chevrolet Buick GMC is one of these customers. Jared King, General Manager of the King dealerships, notes that they have switched to CarGurus SEM Plus because of the performance of CarGurus' listings platform. King adds that SEM Plus has provided 100% increase in sales and leads for the King dealerships as compared to their prior SEM provider. And together, the CarGurus listings and SEM Plus package have become our #1 source of leads and driving sales. We continue to believe SEM Plus will serve as a powerful complement to our listings product in driving high-quality connections to dealers, ultimately, helping CarGurus increase dealer retention and average annual revenue per subscribing dealer or AARSD over the long term.

We've also introduced a regional and nationwide shipping feature for select dealers. The regional rationale for this was consumer centric. We strive to maximize the relevant inventory available to CarGurus' consumers for searches with limited local inventory of certain makes and models. With our nationwide shipping feature, U.S. dealers are able to display their cars to consumers in other parts of the country and transport those cars directly to the consumer. Although this feature remains in relative infancy, we believe it provides a unique value proposition for dealers, allowing a given dealer to greatly increase its addressable geographic market while simultaneously improving the customer experience by expanding available car inventory options in lower inventory markets. Ultimately, nationwide shipping provides another avenue to expand our relationship with dealers and drive AARSD growth over time.

In addition to product innovation, we continued to make investments to increase dealer awareness to bring dealers on our site. Once partnered with us, we aim to retain them by continually growing our large, engaged audience and expanding our high ROI product portfolio. We enjoyed another solid quarter of paying dealer additions with our total global customer base approaching 30,000 total paying dealers. Our marketplace provides value to dealers of all sizes, whether it's large franchise groups or an emerging independent dealer, demonstrating both the scalability and flexibility of our platform. We believe this will continue to attract more paying dealers to our site over time.

In order to execute well and further audience growth and attractive dealer products, we continue to invest in the best possible team to capitalize on the opportunity in front of us. Upon becoming a public company, we set out to add more independent directors with relevant skill sets to bolster our Board of Directors. To that end, we made 2 important additions to our board in the recent months. First, we announced in June we added Steve Conine to our Board of Directors. Steve co-founded Wayfair in 2002, where he helped build the company into a global consumer brand. We're excited to benefit from Steve's engineering and product expertise, and we are looking forward to his contributions.

Second, I'm pleased to announce that we recently added Lori Hickok to our Board of Directors. Lori is a 30-year veteran of Scripps Networks Interactive, having served most recently as the company's Chief Financial and Development Officer. Lori was integral in establishing and executing Scripps' overall growth strategy, and she played a key role in Scripps' talent management, investment prioritization and M&A efforts. Lori will serve as the Chair of our Audit Committee, and we believe her financial acumen will be invaluable as we continue to grow our platform.

We also announced today that Simon Rothman recently departed our Board of Directors. Simon has made significant contributions to our board and the company overall since CarGurus' earliest days in 2006, and we thank him for his dedicated service.

In summary, the second quarter yielded continued momentum across our business, and we are well positioned to execute on our key initiatives in the back half of the year. We will continue to invest in our brand and our products to ensure we deliver the best consumer experience while driving meaningful return on investments for dealers. Above all else, we remain committed to building the world's most trusted and transparent automotive marketplace.

With that, I'll turn the call over to Jason to provide more details on our financial results.

Thank you, Langley. I'll begin with the detailed overview of our strong second quarter performance and then provide our guidance for the third quarter and updated view for the full year 2018.

Starting with the income statement, total revenue was $110.3 million in the second quarter, up 45% year-over-year and roughly $6 million above the high end of our prior guidance range. Marketplace subscriptions once again served as the primary source of the revenue outperformance, growing 44% year-over-year to $97.7 million.

Advertising and other revenue represented the remaining $12.6 million of total revenue, an increase of 49% year-over-year. We view the performance in advertising as a reflection of our increased investments in building brand awareness to augment our large engaged consumer audience, driving greater ROI for our advertising partners. That said, on a quarter-to-quarter basis, we expect advertising and other revenue to exhibit less linear growth than marketplace subscriptions revenue since campaign timing and promotional activity from our OEM and other auto-related advertising partners can vary quarter-to-quarter.

Looking at our performance by geographic segments. U.S. revenue was $106.4 million, up 44% year-over-year, and international revenue totaled $3.9 million, up 87% versus the year-ago quarter. As Langley referenced, we launched our site in Spain in Q2, and we continue to make strides in Canada, the U.K., Germany and Italy.

Turning to our paying dealer count. We ended the second quarter with 29,969 total paying dealers around the world, up 20% year-over-year and representing an increase of 943 paying dealers from the end of the first quarter. Our launch in Spain brings our total global addressable dealer count to roughly 90,000, giving us a sizable opportunity to increase our paying dealer penetration moving forward.

In the U.S., we ended the second quarter with 26,871 total paying dealers, up 15% year-over-year and representing a 610 paying dealer increase from the end of the first quarter, a total net add that was in line with our expectations.

Our inside sales team continues to make great strides acquiring paying dealers, while we continue to focus on maximizing dealer retention via product improvements and through value adding engagement with our dealer relations and customer success teams. Further, we will continue to make targeted investments in our field sales team in certain U.S. regions and international markets to further expand our paid dealer count as well as developing new methods in demonstrating our platform's value such as ROI-driven packaging and pricing that better reflects the size and quality of the audience we're driving to our paying dealers.

Total international paying dealers ended the quarter at 3,098, an increase of 333 from Q1 2018. International dealer growth will likely be nonlinear on a quarter-to-quarter basis as each international market we operate in is at a different stage of development and as we continue to invest heavily in bringing non-paying dealers on to our platform to deliver the best consumer experience in each international market.

The second quarter was a stand-out period for AARSD, as dealers continued to achieve superior ROI from the volume and quality of connections they received from our platform. We saw positive momentum across several areas that drive AARSD growth, principally, connection volume growth in Q2. In the U.S., AARSD was 13,130 at the end of the second quarter, an increase of 19% year-over-year, while international AARSD was 5,037 at the end of the second quarter, an increase of 2% year-over-year.

Given that we are in the early stages of building out our platform in multiple international markets, our primary near-term focus remains on growing our international audience. Our fundamental international market development phases are to grow inventory and audience, grow the number of free dealers on our platform and then paid dealer count and finally, expand AARSD over time. As a result, we continue to expect international AARSD to be lumpy as we execute on our investment plans and work through these phases in distinct international markets.

As in previous quarters, I will discuss our expenses and profitability on a non-GAAP basis, which backs out our stock-based compensation expense and other discrete items. As a reminder, we believe looking at non-GAAP expenses provides a better comparison of our operational performance since our stock-based compensation has become more material post IPO, and therefore, year-over-year comparisons of our GAAP expenses and profitability are not yet helpful.

In the second quarter, non-GAAP gross margin was 94.7%. Total non-GAAP operating expenses were $97.4 million, up 48% year-over-year. Non-GAAP sales and marketing expenses represented 82% of our operating expenses in the quarter, in line with the first quarter.

These investments continue to grow our market-leading audience while driving higher connection volumes for our dealers. Longer term, we expect to generate substantial operating leverage via sales and marketing, but for now we will continue to invest aggressively when appropriate to address the large market opportunity before us.

Investments in our products and technology platform will continue to be a focal point for us moving forward. We continue to add engineering headcount to improve our existing products while innovating to expand into adjacent categories. As such, product, technology and development expense grew 96% year-over-year to $9.2 million. These investments are supporting broader feature sets in our existing products, new products and services such as SEM Plus and nationwide shipping and international expansions such as our launch in Spain.

We remain committed to pursuing top line growth while maintaining healthy profitability. To that end, we delivered non-GAAP operating income of $7 million, which was ahead of our guidance range due to the higher revenue in the quarter, offset slightly by increased investments in the business to fuel the company's long-term growth. We will continue to take advantage of top line outperformance by proactively reinvesting in our business.

Non-GAAP diluted earnings per share were $0.06 for the quarter, which was $0.02 above the high end of our guidance range. On a GAAP basis, we delivered GAAP gross margin of 94.6% and total GAAP operating expenses were $102.9 million. Additionally, second quarter GAAP operating income was $1.4 million, down from $6 million in GAAP operating income in the year-ago quarter. The reduction in GAAP operating income is primarily the result of increases in stock-based compensation expense. We recognized $5.6 million in stock-based compensation expense in Q2 2018 compared with less than $100,000 in stock-based compensation expense in Q2 2017. GAAP net income attributable to common shareholders totaled $31.3 million.

During the quarter, we recognized a $29.1 million tax benefit stemming from stock deductions from the taxable benefits of equity-based compensation and recognition of federal and state research and development tax credits. GAAP diluted earnings per share totaled $0.28 in the quarter.

Geographically, our GAAP operating income was $9.7 million in the U.S., and we had a GAAP operating loss of $8.2 million in our international segment.

Turning to our balance sheet. We ended the second quarter with $142 million in cash and short-term investments, roughly in line with our balance at the end of the first quarter. We generated $17.5 million in cash from operations in the second quarter and $16.8 million in non-GAAP free cash flow, which includes capital expenditures and capitalized website development costs of roughly $700,000.

During the quarter, we withheld and remitted $17.5 million in withholding tax payments from RSU share settlement stemming from our equity awards plan, which allowed us to return the corresponding shares to our equity pool. Moving forward, the impact of withholding and remittance of taxes related to our equity awards plan should moderate and have less material impact on our overall cash flow. Overall, we believe our cash generation profile continues to reflect the capital efficiencies of our differentiated technology-driven model.

I'll close my prepared remarks with our outlook for the third quarter and full year 2018. We are raising our full year revenue outlook to a range of $436 million to $438 million, up from $415 million to $418 million previously. We expect the realized increases in our marketplace subscription segment to carry over to the back half of the year while acknowledging we will likely experience non-linear performance in our advertising and other segment.

We are also raising our full year outlook for non-GAAP operating income to a range of $28.5 million to $30.5 million, up from $25 million to $28 million previously. Our outlook accounts for investments to drive long-term growth, including investments focusing on building our brand, driving audience and connections growth and building out our products to further improve our dealer value proposition. In addition, we are increasing our full year non-GAAP EPS guidance to a range of $0.22 to $0.23 per share, up from $0.19 to $0.21 previously.

Focusing on the third quarter, we expect total revenue to be in the range of $112 million to $113 million, non-GAAP operating income in the range of $5.5 million to $6.5 million and non-GAAP EPS of $0.04 to $0.05 per diluted share.

In summary, the company continues to perform well against our financial objectives and execute on our key strategic initiatives for 2018, and our teams are focused on pursuing the large market opportunity before us. Our investment in audience growth and consumer awareness is proving effective. Coupled with ongoing listing enhancements and new products, we continue to drive ROI, dealer satisfaction and resulting expansion among our paying dealers in the U.S. We're leveraging these platform-wide investments to replicate this success in other countries as we continue to invest in our international segment. We're pleased with the progress we've made thus far in 2018, and we look forward to finishing this year strong and setting the stage for continued growth in 2019.

With that, we'll open up the call for Q&A.

(Operator Instructions) Our first question comes from Heath Terry with Goldman Sachs.

Wondering if you could give us a sense, just maybe a little bit more detail on what you're seeing in international markets, particularly as you've launched or gotten further into your launch in the U.K. and Germany, specifically on conversion on free to paid on your user -- or dealership base. And then just on traffic in the quarter, continuing -- obviously, continue to see strong growth in traffic in the quarter. If you can help us just sort of understand a little bit more on the breakdown between organic and paid there, that would be really helpful, particularly in the U.S. given the dealer net add number there.

Heath, it's Sam Zales. I'll take the first question related to international. We're seeing continued scaling on the value proposition for dealers. In the international markets, I think, I've mentioned the same pain points exist for consumers and dealers in the markets we've entered. I think I'd add that's the same with Spain that just launched. For the dealer community, it's looking for a return on investment to their customer acquisition expenditures and channels they're using, and we provide that great opportunity for them. The conversion from basic to paying is continuing to go well. I think you see the 83% growth in our paying dealer base in international markets. That's a sign that, one, when they try the free program as they do freemium models here in the U.S., it really leads to a significant opportunity to grow customer acquisition, moving from the free to paid programs. We've been able to demonstrate that the connection volume growth from the basic program to one of our premium products and the change in conversion rate of selling those vehicles is substantial, and the international markets are behaving the same way the U.S. is, that we're converting a significant number of those dealers to the paid programs. As you know, our opportunity in international markets is to, first, gain inventory on the free program, so getting as much inventory as possible to make our consumer experiences as best it can be and then converting those dealers once we've built the consumer audience to see that benefit in connection volume and close rates. And so we're doing that significantly and at a great pace in those international markets. I'll turn it to Jason on the consumer side.

So yes, the U.S. traffic -- our U.S. traffic in the quarter was -- we were very pleased with it. It was an acceleration, as you've heard. We have talked in the past that we don't break out our channels specifically. At a high level, we look at paid and unpaid, and paid includes social retargeting, search and display. And unpaid includes direct, branded search that comes through search engines but uses our brand and then organic search. I think one of the bigger drivers that we're seeing now is the influence of our brand campaigns in the U.S., specifically our TV campaign. And that typically takes time to build momentum, and we're seeing that now. We're seeing the confluence of having been in the market since midyear last year and continuing at a steady state there as well as the high season for car shopping kicking in, in Q2. And so we are seeing disproportionate growth in direct traffic and in branded search, but that said, we're seeing growth across the board in all of the traffic channels that we're in. We have spoken a lot about our growing spend in brand, but we should also mention that our algorithmic traffic acquisition team is a highly sophisticated team that we think is market leading. But they're always making efficiency improvements that are allowing them to scale on the same or less spend as well. And so we're seeing growth across the board. The brand is the biggest new factor within the last year for sure.

Our next question comes from Ralph Schackart with William Blair.

Two questions if I could. First, on domestic AARSD growth of around 19%, which I think is about the strongest growth since the first quarter of '16, at least with the data that we have in our model. Jason, I know you talked about connection volume growth having a strong impact in the quarter but just curious if you could just take a step back and sort of assess how are you able to sort of drive that metric at the fastest rate in nearly 2 years given the scale that you're at today. That's my first question, and then I have a follow-up.

Sure. So AARSD, as I think you guys have learned in your models, is a complex metric in terms of what builds it up. It's driven by, really, 3 levers, which we've outlined: audience, pricing and new products. In this past quarter and the past couple of quarters, we've had strong audience growth, and strong audience and connection growth tends to lead to stronger renewals. And that has historically been our largest -- our strongest lever of growing AARSD. And so that continues to be our strongest lever, followed by new products and unit pricing. We've always said that unit pricing is going to be a very slow, methodical march for us, and so I would remind everyone that, that's not going to be a lever that we pull hard anytime soon or maybe ever. So the order of those 3 levers has remained. I think one thing that is an overriding factor to all of that, though, as it relates to AARSD is just the math of how AARSD works. And we are, for the calculation, replacing an old quarter, 5 quarters ago, with a new quarter, and that -- this most recent quarter has the benefit of renewals for the 3 quarters in between. And we've done a good job with renewals lately. And so as our revenue grows faster than our dealer count grows, we're going to see AARSD grow. So some of it's the math. But really, the headline is our connection growth and our audience growth has continued to be really, really strong, and that's helping with renewals.

Yes. Ralph, this is Langley. One thing I would probably add to Jason's comment is that I think the thing we are excited about probably going forward is the fact that we are now kind of building a suite of products in addition to listings. Obviously, we had our Dealer Display product that we launched over a year ago and that our search engine marketing product has gotten quite favorable reviews from the dealer community. It's still early days on that, but we're -- we feel like we're layering on multiple products and kind of building a suite so that it's not just about our listings product but we kind of have a whole suite for dealers. And as Jason said, that kind of adds to the velocity of AARSD in the future.

It's actually a good segue to my second question on the SEM Plus product. Sounds like you're getting good initial feedback. Maybe if you could help us understand or frame the revenue opportunity for that product as you see it today. Give us a sense of how you see that product rolling out and progressing. And any compares, if it's relevant, to how Dealer Display rolled out versus the signals you're seeing with SEM Plus would be helpful.

Yes. I'll let Jason talk probably to the market sizing. But I think the important point that we've noticed in the marketplace is that the technology that we bring to bear is significantly different than what most dealers have been dealing with today. They kind of -- they tend to fall into 3 camps. One, they're either trying to do it themselves, which is something we can pretty easily overcome in terms of success. The second one is they might be working with a local agency, and typically those sort of agencies tend to be subscale in terms of the volume of keywords they can provide to a dealer. And then thirdly, there might be a larger digital marketing agency they might be working with. I think the key that we found so far is that the sheer scale of the keywords that we can bring to bear for a given dealer allows them to accomplish kind of really 2 things. One is volume. Typically, the architecture that we built over the last year has just tested as many as 1 billion keywords. And when we typically talk to a new dealer, we'll ask them how many keywords have you been working with, with your previous solution. And we often get responses of 250 or 300. And typically, when we light up a new account, we're talking about 20,000 keywords. And that really has, as I said, kind of 2 significant levers for that dealer. Number one is scale. We can drive a lot more traffic; and secondly, cost because we're -- the architecture we built is really about mining the long tail. It's going for the many, many keywords that most people have overlooked instead of going for the obvious ones, which are ironically more expensive. So it gives them more scale. It also gives them a lower cost per VDP and cost per downstream activity. So we're really excited about what we've seen in the marketplace so far. And Jason can maybe talk a little bit about the kind of market sizing.

Sure. So I'll do a quick top-down, Ralph. The -- we believe that auto dealers spend several billion in searches, in the sort of $4 billion to $6 billion range and -- in gross spend. And so if you think about the fees associated with that, if it were to all be outsourced, which is not because a lot of dealers will do it internally, then you're looking at, call it, $1 billion in round numbers. So that $1 billion of potential fees, but again, some of it's in-sourced, is really fragmented. There's a couple few businesses that are in -- maybe approaching $100 million net revenue range, and then there's a lot of businesses, we've referred to them as sort of local agencies to the dealers, that are in the sub $50 million, sub-$30 million net range. So it tends to skew toward franchise dealers. It tends to skew toward larger dealers. And so we're being smart about who we target for this. This is a rip and replace as you've heard us say, so it's a longer sales cycle. It's -- on an account level basis, it takes longer -- it would take longer to ramp and display. The flip side is it's a higher ticket item than our display product. And so we see an opportunity to build a business here in our installed base that can really help accelerate our growth or contribute to our growth in the next several years when we reach meaningful penetration of our paying dealers, paying listing dealers.

Our next question comes from Ron Josey with JMP Securities.

One question and just a quick follow-up. Great quarter. So just on the new product. Langley, you mentioned about regional and shifting feature. And so you said it was nationwide shipping for select dealers, and you mentioned select dealers specifically. So just wondering, what does a dealer have to do to qualify for this feature? And then what is CarGurus, what are you all responsible for with this product? Are you planning the logistics and the shipping? Or are you actually taking the shipping on yourself? I'm wondering only because this might broaden your market significantly. And also given users are buying these cars sight unseen, wondering if there's a return policy. And then, Jason, you mentioned on the AARSD question just now replacing an old quarter with new quarters that have -- and you've done a better job on renewals. Just remind us on the better job you've done on renewals, what you've done to improve that.

Ron, it's Sam Zales. I'll take the question related to what we call nationwide shipping. First, I'll echo a couple comments that were made in the opening statements Langley made. We developed this offering at the understanding of consumer experience. So we found in certain geographies around the country, remote locations where a consumer was searching for vehicles and found a limited set in that particular region and hope to provide the biggest search opportunity with the best dealer experience for those consumers. And we said, at a certain search level, search criterion provided, not a significant number of searches, we could go to dealers in a surrounding area that can shift to that consumer directly and is already doing so in their general practice of their business and can do so very effectively. We could provide those searches of those vehicles outside of the area that could be shipped directly at no cost to those consumers. And so we built the product at a consumer experience base and went to dealers that already have the capability. We do not manage the logistics. I want to be clear about that. It is dealers who are already able to ship that vehicle to the consumer directly at a really positive consumer experience. So we built this with a 250-mile radius today and said that, that consumer -- if the dealer is shipping a vehicle to that consumer, we will allow that dealer's vehicles to appear in those search results, and we will be very clear to that consumer to say that the vehicle will be shipped to you. The price is included. Shipping fee is included or there's no shipping fee, and we let them know that it will be delivered to their doorstep. As with many companies in this industry, you know of the Vrooms and the Carvanas. They have a return policy. Our partners that we work with today will honor a similar one. So the consumer, if they're dissatisfied with the experience, is able to return the vehicle. And we've seen a tremendous consumer-positive experience in that regard of highly -- high ratings for that dealer experience. And the dealers themselves, getting the opportunity to put their vehicles in front of a broader consumer audience, that's great for their business, helps them close more business and helps CarGurus create a bigger subscription revenue stream from those dealers.

Sam, if I could follow up real quick. Sorry to interrupt, Jason. Just do you know -- do you have a sense of your current traffic, how many folks actually look in 250-plus mile areas to begin with and are willing to travel to pick up the car themselves? Does that make sense?

It does, Ron. I don't have the number in front of me. There are consumers that have the flexibility, obviously, with our platform to change the radius of their search and are willing to go and shop, and we've seen this in international markets as well to drive to the dealership at a farther location. We thought this particular consumer experience would provide both dealer and consumer with an optimized process of saying, "I don't need to leave and drive 250 miles. The dealer is going to ship it to me directly." And that's been a big boon to both sides of the marketplace.

So I'll start to address your renewals question, and then Sam may add to it. This is one of Sam's [grooves], so I'll give him a second to breathe. So yes, I would say there are a few factors driving why I characterized us as getting better at it. The first is that we are becoming increasingly ROI driven, and that's part of our messaging with the dealer. It's how we are trying to help educate the dealers. You've heard us talk about our ROI calculator, and it sounds like a very simple thing, but that's really an education process that we've been taking them through for several quarters now and I think that's starting to resonate with them. And that allows us to focus on the gross profit that we're driving to their dealership. And the -- on a relative basis, low cost that they're paying us to achieve that gross profit. And that helps change the tenor of the discussion from "How could you increase me by X percent?" to, "Wow, I'm [not] sure every dollar I'm spending I'm getting X back." The second thing is I think our account teams are being more consultative. They're really partnering with the dealers to help them merchandise their cars better and understand the market drivers better and frankly, just be more successful on our marketplace. I think the third factor is our scale is more obvious to dealers. There's more evidence from a variety of factors -- or from a variety of sources rather that just display the facts that we are the leader and in the last few quarters have become the dominant leader. I think that helps the conversation. And then the last is we're sort of going back to our roots and using data to really predict dealer success but also dealer risk and identifying those leaders that are not seeing the success that we'd like or are behaving in ways that we don't think could lead to success for them, and we're being more proactive in addressing them in real time rather than waiting for any sort of historical renewal cycle.

Our next question comes from Tom White with D.A. Davidson.

Just on the guidance, so you're raising the full year revenue guide kind of well in excess of the 2Q beat. Just kind of curious what are the specific drivers of the improved kind of implied outlook for the second half? Is it some of the AARSD strength from maybe new products? Is it the -- maybe the increased pace of dealer addition from some of the investment and field sales? And then just on audience, obviously, audience growth is spectacular. You talk about connections growth. Any sense you can kind of quantify what the kind of down funnel connections growth is that compares to the unique user growth?

Sure. Tom, it's Jason. So on the guidance. We have had really a strong first half of the year. We believe we'll have ongoing strength in subscriptions, and so we have flowed through our first half strength through the second half of the year. We do acknowledge, as you've heard us say, that advertising revenue is based on campaign activity and promotion timing from OEMs and others. And so that's less predictable. So we've -- our guidance raise has been, I would say, principally on the backs of the first half strength and the belief that our audience growth that we've been seeing now is impacting renewals that we're doing now, which will support strength in the second half. As far as uniques and audience and how they relate to connections, it's a good observation. I mean, there is not 100% correlation by any means to -- between connection volume and audience growth. And in fact, as I've said, one of our biggest -- one of the biggest new influences in our business is brand spend in Q2 this year versus Q2 last year. And as I'm sure you see in other business, television-based or mass market media based brand advertising tends to attract to higher funnel shopper than search does. And so those are likely going to be -- not likely, but they're often unique to our less ready to convert right away. You've seen that our sessions per unique has stayed fairly stable, and so that hasn't had a huge influence on it yet. But it is changing the shape a little bit of our audience, and as a result, we're not going to have lockstep audience growth and connection growth. The other thing, too, is that not all connections are created equal, and dealers value certain types of connections more than others, and user behavior changes over time. It changes with mobile platforms, and it changes with site changes that we may make. It changes with how dealers might be encouraging users to connect with them and how they respond to different types of connections. So there are several factors that go into it. Our goal is to not necessarily just maximize connections, but it's to maximize the right kind of connection. So we think about it as quantity and quality, not just quantity. But suffice to say, with such strong audience growth, we are seeing really strong connection growth in that supporting dealer ROI and renewals.

Our next question comes from Aaron Kessler with Raymond James.

Couple of questions. First, you may have mentioned it before but just if you can provide us an update on kind of latest thoughts on your price testing and how that may evolve here. And second, just maybe your general philosophy in kind of reinvesting upside on revenues looks [good]. Obviously, good upside in Q2. I thought some of that was reinvested kind of across operating expenses and then how we should think about maybe out leverage over kind of the near to intermediate term as well.

Aaron, it's Sam Zales. I'll take price testing and turn it to Jason on revenue. We -- I think I mentioned maybe a quarter or 2 ago, we were in the first inning of price testing. We -- so maybe we're in the bottom of the first and the top of the second. We have had great success thus far in test mode of looking at different pricing and packaging arrangements. I think Jason hit it well on the last question that we're starting with an ROI-based analysis of all of our renewals, pricing of all of our new product pricing and understanding the ROI we're driving to dealers. And that as foundation is very different than our history of looking at things like a lead-based pricing model, and so we're getting much more sophisticated about our understanding of the impact of the connections we're driving and the value we're driving to dealers. We are testing things, as I had mentioned, as auction-based models, pay-per-performance models. And we're still in the early innings. You're seeing the success of some of that reflected in the AARSD growth, but it's too early to say that we have a perfected model of 1 or several of those that will be used permanently as we move forward. I think this is going to be an iterative process, where we continue to find that incremental growth to our pricing and packaging strategies and will hopefully reflect that in ongoing AARSD growth for the business. If we keep driving the connection growth, we know we're driving ROI and we can keep testing models that reflect a better balance of the value we're driving to dealers versus the price they pay for the subscription.

Yes. Aaron, it's Langley. I'm going to -- I'll take the one on reinvestment. So I mean, I think Jason alluded to it in his earlier remarks that our focus on reinvestment when we over-deliver on revenue is to really focus on probably 2 things. Number one is obviously technology. We believe that our core competitive advantage against all our competitors is, in fact, technology. So we're building up our tech team as kind of as fast as we can to push new products into the marketplace, but up and above that and probably at least in the near term, more significantly is really brand. Our consumer marketing team has done a great job of digging in and exploring kind of off-line TV -- more TV than radio, although there's going to be some radio testing as well but principally, TV branding. And their early returns have been really encouraging in terms of both aided and unaided awareness. We feel like we have a real point of differentiation from a product standpoint against our competitors, and so we feel this is the best time to stake our claim in the marketplace and reinvest whatever dollars we have where we over-perform into building that brand.

Our next question comes from Dan Kurnos with The Benchmark Company.

Jason, I don't want to mischaracterize what you said earlier, but just going back to the dealer add side, as you kind of go through the renewal process, pace of adds domestically, I think people are kind of sort of wondering sort of where the saturation point eventually occurs. And while that's still probably a fair ways off, it sounds like maybe not as aggressive with some intentional decisions on your part either not to renew or not to push the bad actors, as you put it. Is that a fair characterization? And then on the international side, as you guys really start to get a lot more penetrated into the marketplace, more brand awareness, just love to get your sense on either how competitors are reacting to you guys thus far when you launch in new market like this now that you're more of a known quantity, how quick people are to react to the marketplace internationally.

Sure. Can you -- the first part of your question, I'm happy to take. What was the reference to bad actors, though? I missed that.

Yes. So I think the concept was on sort of renewals as you guys go through the process, I think you referenced earlier. And maybe I'm mistaken in sort of your comments. But just in terms of the dealers that you don't think are necessarily good actors or they're doing well with the data or -- how would you characterize it earlier? I'm just trying to get a sense that there's sort of intentional, I don't want to call it churn, but I guess, it's kind of really what it is, if you guys are growing the U.S. dealer count, if that's kind of limiting maybe pace of adds over the near term as you go through the renewal process.

Got it. Okay. So yes, I think I -- if I used the term bad actors, I don't remember using that, but we certainly don't necessarily think about it as that. We think about it more as that we are getting to a place where we're delivering the scale and the quality that is leading the market and that we should be compensated in a way that is still a resoundingly positive ROI for the dealer but that is in line with what the market is supporting for that sort of pricing. And so we do -- I guess, walk away from dealers who are not willing to pay what we believe are fair market prices and still, frankly, more attractive unit pricing than a lot of other options out there. It's not so much that they're a bad actor per se in terms of having various activity. In terms of the new dealer adds, the first thing I would say is we think about our dealer -- our available dealer pool right now as about 90,000 dealers across all the markets that we're in, and so we're just bumping up against 30,000 of 90-plus thousand, sort of 90,000 to 100,000 in all of our markets. So with that measuring stick, we're at 1/3 or less than 1/3 of penetration, and we feel that we have a long way to go. On a market-by-market basis, we look at it by segments, where we still see a lot of opportunity. But in an aggregate level in the U.S., we do expect and have signaled that we are going to see a percentage growth slowdown as we achieve more market penetration. If you look at our -- some of our larger competitors, they're in the, call it, 20,000, 22,000 paying dealer range, and we're up at much higher than that. And so we always refer to it as uncharted territory, and we think we're -- because of our business model differences, we appeal to a larger set of dealers. How far we get into that 45,000 or so in the U.S. is -- remains to be seen. We're targeting to still exceed 30,000, and we think we'll get there. But we're not going to sort of steam there at the same pace and then just stop. It's going to moderate as we go. But I think the key theme for us is how do we build products at a platform level that can work for the 90,000 to 100,000 dealers and soon more as we ultimately launch more countries. And in the meanwhile, we are going to expect to be paid a market rate. And so we have a freemium product, and dealers are welcome to stay on that freemium product. But if they're going to enjoy the benefits of paying, then they will need to pay sort of reasonable fares for that.

Yes. Ben (sic) [Dan], it's Langley. So I could take your second question about -- I think you asked more about international reaction to our product in those markets. We have seen in Canada and the U.K. the incumbents kind of react to our transparency. And to take a page out of my past history when I was at TripAdvisor, we saw some of our competitors and actually even customers start putting hotel reviews on their website when TripAdvisor became so popular. And I'll kind of cut to the punchline. Our competitors actually started putting user reviews up, but they would put positive reviews. They wouldn't put any negative reviews because they were worried about merchants being upset. And so some of our competitors are doing that. They're putting -- when a car is a great deal, they'll kind of highlight that on the VDP, but they won't talk about when a car is a bad deal. My reaction to that is consumers aren't stupid. They'll see through when you're doing semitransparency, is kind of what I would call it. Some of our other competitors are taking even more seriously, and they are showing both good and bad deal analysis on the VDPs. But more importantly, and I think this is where it cuts to perhaps their business model, they're unwilling to change their search order to rank the cars, where the best deals show up at the top of the search results. And the analogy I would give on that one is when Yahoo! was out in the marketplace with Inktomi and they were making all that money on paid inclusion, and then Google came along with Google PageRank and said, "Well, no, I'm not going to sort the links based on who pays me the most money. I'm going to sort them based on this mathematical algorithm." Anyhow, fast forward to today, Google destroyed Yahoo! because of the fact they were using an unbiased mathematical model. And consumers saw that, and they flocked to Google. We are seeing that in both -- well, we've seen it in the U.S. market. We're seeing it in international markets where consumers are smart, and they'll see when they -- they'll flock towards full transparency. And our competitors are in a bit of a box because, let's be clear, they built some really formidable great businesses, but they've done it basically on a paid inclusion model, where they've -- the sort order is based on who pays them the most money. And they're in a dilemma because they can't change that sort order without blowing up their income statement, and some of these competitors have $300 million -- $350 million revenue lines they have to worry about. So we believe that we have a fundamental advantage because of our product focus and how we think about technology and how we think about the consumer and how we sort inventory. And I think, yes, they are reacting to our entry, but they're doing it in, again, as I referred to, as a semitransparent way, and I think consumers will see through that pretty quickly.

Our next question comes from Mark Mahaney with RBC Capital Markets.

I just want to ask a blue sky question. If we think about 90,000 dealers or 100,000 dealers worldwide and long term, what that -- if you put the North American or the U.S. AARSD on that, that's a pretty big market. So just talk about the structural differences between the AARSD you're seeing in the U.S. and then international markets. And I know the international AARSD you have now is a lot lower than the U.S. AARSD, but it's also blending in a lot of countries. The curves that you're seeing there in AARSD growth in Canada and U.K., the oldest international markets, do they look like what you've seen in the U.S.? Are there any structural reasons why long-term international AARSD shouldn't rise up closer -- much closer to the U.S. levels than it is today?

I can start this one. So Mark, the -- first of all, the 90,000 is the markets that we're in right now, so if you -- so obviously, that becomes bigger if and when we enter other countries. This -- in terms of why it's different, for us right now I would say the biggest driver to that is just our size in those markets. And so if you think about it on a connection per dealer basis, we're just driving far fewer connections per dealer in those markets than we are in the U.S. right now. And so if you look back at our U.S. business, connections per dealer has grown considerably. There's a period when you're adding both dealers and adding connections, growing connections such that a connection per dealer metric becomes harder to grow, but once you onboarded a good portion of dealers free or paid and then you just start continuing to grow connections, then you see that metric go up. There are some -- there are 2 structural things that I would say. The first is that different markets just have different unit pricing in the market, and some are higher than the U.S. Some are lower than the U.S. And I think that's a function of the dealer P&L models, including the sort of dealer and OEM relationship infrastructure. But -- so that's something that I think will drive the ultimate potential of a cost per connect or revenue per connection in our eyes. I think the second thing is consumer behavior. So in some markets, consumers are more inclined or willing to use a listing site to connect with a dealer, and other times, they're less inclined. And in those markets where they're less inclined, providers -- marketplace providers need to think of more of an advertising like or an advertising heavy model. So there's a couple of structural things, but I would say, first and foremost, it's our maturity and size of the market.

Ladies and gentlemen, there are no further questions at this time. I will now turn this call back to management for closing remarks. Thank you.

We didn't have anything to add, but we certainly appreciate all of the input, all of the interest, the great questions, and thanks, everyone, for your time.

Thank you. This concludes today's conference. All parties may disconnect. Have a great day.